Activation of the hedgehog pathway in advanced prostate cancer

Abstract

Background: The hedgehog pathway plays a critical role in the development of prostate. However, the role of the hedgehog pathway in prostate cancer is not clear. Prostate cancer is the second most prevalent cause of cancer death in American men. Therefore, identification of novel therapeutic targets for prostate cancer has significant clinical implications.

Results: Here we report that activation of the hedgehog pathway occurs frequently in advanced human prostate cancer. We find that high levels of hedgehog target genes, PTCH1 and hedgehog-interacting protein (HIP), are detected in over 70% of prostate tumors with Gleason scores 8-10, but in only 22% of tumors with Gleason scores 3-6. Furthermore, four available metastatic tumors all have high expression of PTCH1 and HIP. To identify the mechanism of the hedgehog signaling activation, we examine expression of Su(Fu) protein, a negative regulator of the hedgehog pathway. We find that Su(Fu) protein is undetectable in 11 of 27 PTCH1 positive tumors, two of them contain somatic loss-of-function mutations of Su(Fu). Furthermore, expression of sonic hedgehog protein is detected in majority of PTCH1 positive tumors (24 out of 27). High levels of hedgehog target genes are also detected in four prostate cancer cell lines (TSU, DU145, LN-Cap and PC3). We demonstrate that inhibition of hedgehog signaling by smoothened antagonist, cyclopamine, suppresses hedgehog signaling, down-regulates cell invasiveness and induces apoptosis. In addition, cancer cells expressing Gli1 under the CMV promoter are resistant to cyclopamine-mediated apoptosis. All these data suggest a significant role of the hedgehog pathway for cellular functions of prostate cancer cells.

Conclusion: Our data indicate that activation of the hedgehog pathway, through loss of Su(Fu) or overexpression of sonic hedgehog, may involve tumor progression and metastases of prostate cancer. Thus, targeted inhibition of hedgehog signaling may have significant implications of prostate cancer therapeutics.

Figure 1

Detection of PTCH1 expression in prostate cancers. Protein expression of PTCH1 was detected by immunostaining. PTCH1 antibodies (Santa Cruz Biotechnology Cat# 9149) were tested in Ptch1^-/- null MEF cells (A). While Ptch1^-/- null MEF cells had no positive fluorescent staining with PTCH1 antibodies, transfection of PTCH1 expressing plasmid lead to positive staining (green, indicated by an arrow, 400×). Immunohistochemistry of prostate cancer specimens with PTCH1 gave negative (B-a, 200×) or positive (Red in B-b, 200×) signals. When PTCH1 antibodies were pre-incubated with the very peptide for raising the antibodies, no positive signals could be observed (B-c).

Figure 3

Detection of HIP in human cancer specimens. By Western blotting, HIP antibodies (R&D systems Cat# AF1568) recognized one band between 75 and100 KD (A). Expression of endogenous HIP was detected in two GI cancer tissues, which were known to contain activated hedgehog signaling (data not shown here), but not in the matched normal tissue (B). Immunohistostaining of HIP I prostate cancer showed a similar pattern to PSA (C, 200×)

Figure 4

Detection of Su(Fu) in prostate cancer specimens. Su(Fu) antibodies (Santa Cruz Biotechnology Cat# 10933) recognized only one single band (54-Kd) in D283 cells (A). Following treatment of a specific SiRNA of Su(Fu), the endogenous Su(Fu) band was greatly reduced (B). Immunohistostaining with Su(Fu) antibodies in prostate cancer specimens revealed positive (C, in red, 200×), negative (D, 200×) or weak staining (E, red, 200×).

Figure 5

Inactivation of Su(Fu) in prostate cancer. Two TURP (Transurethral resection of the prostate) tumors with loss of Su(Fu) expression were confirmed by Western blotting (A). One mutation of Su(Fu) found in prostate cancer PC48 is shown in B, which is predicted to create a STOP codon in the Su(Fu) coding sequence +1318. The levels of Gli1 and PTCH1 transcripts in prostate tissues were detected by real-time PCR (see methods for details) (C). Tumor tissues had higher levels of the target gene transcripts.

Figure 2

Co-expression of PTCH1, PSA and Shh in prostate cancer specimens. Immunohistochemistry of prostate cancer specimens with PSA was used to confirm the cancer region. Positive staining was in red. Positive staining patterns of PTCH1 and Shh antibodies (Santa Cruz Biotechnology Cat# 9024) were similar to that of PSA staining. PC23 (A-C) was from tumors with Gleason score 7 (200×). PC38 (D-F) was a tumor from Gleason score 10 (400×) (see Table 1, Additional file 1 for details).

Figure 6

Cellular functions of the hedgehog pathway in prostate cancer cells. Expression of hedgehog target genes, PTCH1 and Gli1, were detected by real-time PCR (A). DNA synthesis was detected by BrdU labeling (B). Over 1000 cells were counted under fluorescent microscope for the percentage of BrdU positive cells, and the experiment was repeated twice (C).

Figure 7

Effects of cyclopamine on cell invasiveness of prostate cancer cells. Cell invasion assay of prostate cancer cells was performed using BD Bio-coat cell invasion chambers (A). The rate of cell invasion was calculated by dividing cell numbers penetrated the matrigels by the number of cell in the control chambers (without matrigels) (B).

Figure 8

Cyclopamine induces apoptosis in prostate cancer cells. Cyclopamine-mediated apoptosis in prostate cancer cells was analyzed by TUNEL assay. TUNEL positive cells were indicated by arrowheads. Cells with expression of Gli1 under the CMV promoter (indicated by the arrows) did not undergo apoptosis (n = 500).